High-Resolution Fluorescence In Situ Hybridization to Detect mRNAs in Neuronal Compartments In Vitro and In Vivo

Abstract

The localization of specific mRNAs into dendrites and/or axons is an important mechanism to enrich ­proteins at their sites of function and influence neuronal development, plasticity, and repair. The fluorescence in situ hybridization (FISH) methods described here have provided high sensitivity and resolution enabling investigation into the mechanism, regulation, and function of mRNA localization in vitro and in vivo. Two methods are described in detail. The first method employs digoxigenin- or fluorophore-conjugated oligonucleotide probes for the detection of localized mRNAs in dendrites, spines, axons, and growth cones of cultured neurons. The second method employs digoxigenin-labeled RNA probes and fluorescence tyramide amplification for the detection of less abundant mRNAs localized to dendrites in vivo. Both methods enable the visualization and quantification of mRNA granules, and changes in their localization in response to various stimuli. The high-resolution FISH technology described here has broader applications beyond the study of mRNA localization. It enables the quantitative analyses of developmental and cell type-specific patterns of gene expression, and how these are modified by physiological signals or during disease states.

1 Introduction

Post-transcriptional gene regulation through mRNA localization, translation, and degradation plays an important role in neuronal development, structure, and function. The analysis of steady state and stimulus-induced mRNA expression within distinct neuronal compartments is critical for understanding neuronal cell biology. Early experiments used radiolabeled probes to discover the presence of abundant mRNA molecules in dendrites in vivo and in vitro, such as microtubule associated protein 2 (MAP2, (1)) and CaMKIIα mRNA (2) in brain sections, and MAP2 mRNA in dendrites of cultured neurons (3). The development of new nonisotopic in situ hybridization methods using digoxigenin-labeled probes detected by fluorophore- or alkaline phosphatase-conjugated antibodies revealed the dendritic and/or axonal localization of mRNAs, in vitro (4–6) and in vivo (7), that had previously appeared confined to the cell body (8).

Fluorescence in situ hybridization (FISH) allows for high-­resolution and quantitative analysis of mRNA localization within distinct subcellular compartments, such as RNA granules within axonal growth cones and dendritic spines (5, 6). This FISH method was first used to detect poly(A) mRNA in processes of cultured neurons (8). Since then, a number of technical improvements to our methodology have further enhanced sensitivity, reduced nonspecific background, and enabled quantitative analysis of digital images (5, 6, 9–11). Here, we describe our current methodology using fluorophore-conjugated antibodies to detect digoxigenin-labeled oligonucleotide probes (Fig. 1). More recently, the use of fluorophore-labeled oligonucleotide probes has enabled multiplex detection of several mRNAs in non-neuronal cells (12). Here, we describe our method using fluorophore-labeled oligonucleotide probes to visualize mRNAs within dendrites and spines of cultured neurons (Fig. 2). Both FISH methods used with high-resolution widefield microscopy and image deconvolution permit detailed analysis of mRNA granules in specialized neuronal compartments such as dendritic spines and axonal growth cones.

Fig. 1.

FISH detection of poly(A) mRNA in dendrites and axons of cultured hippocampal neurons using digoxigenin-labeled oligonucleotide probes. (a) A specific poly(A) mRNA signal is detected in the cell body and dendrites of a 16 DIV neuron (DIC overlay, top left    ). The poly(A) mRNA signal maintains high signal-to-noise ratio in distal dendrites (>100 μm from the cell body, top center  ). 14 DIV neurons were transfected with RFP (top right  ), and can be visualized in tandem with poly(A) mRNA (bottom left    ). Concurrent synapsin immunostaining shows poly(A) mRNA granule localization at ­synapses (bottom center  ). (b) Poly(A) mRNA granules are detected in dendritic spines (top left, arrowhead  ). The spine is filled with RFP signal (top right  ), which overlaps with poly(A) mRNA (bottom left  ) and synapsin (bottom right  ) signals. (c) Poly(A) mRNA is detected in the cell body and neurites of a 3 DIV neuron (DIC overlay, top left  ). The poly(A) mRNA signal extends the length of the axon (top right  ) and into the growth cone palm and filopodia (arrowheads, bottom panels).

Fig. 2.

FISH detection of CaMKIIα mRNA in dendrites and spines of cultured hippocampal neurons using fluorophore-conjugated oligonucleotide probes. (a) FISH with an antisense Cy3-labeled probe specifically detects CaMKIIα mRNA as compared to the sense CaMKIIα probe. CaMKIIα mRNA is detected in the cell body and through the dendritic arbor as seen in the DIC overlay (top right). (b) The antisense CaMKIIα probe is sufficient to detect mRNA molecules in dendritic spines (arrowheads). These images are magnified from the boxed area in (a).

Although several isotopic and nonisotopic in situ hybridization methods have provided sufficient sensitivity to detect abundant mRNAs in well-defined dendritic laminas of the brain, such as in the hippocampus and the cerebellum (13), they are not suitable to detect less abundant mRNAs in dendrites, and also lack the subcellular resolution to visualize dendritic mRNAs in brain areas that do not have a defined laminar structure (e.g., cortex or midbrain). Here, we describe our FISH method for the subcellular analysis of mRNA distribution in vivo. To optimize sensitivity, this method utilizes digoxigenin-labeled RNA probes combined with a fluorescence tyramide amplification method. A similar method has been previously used to detect immediate early gene expression, e.g., Arc/Arg3.1 mRNA, in tissue with subcellular resolution (14, 15); although, the presence of Arc/Arg3.1 mRNA within dendrites was not shown. Until now, very few studies have shown low-abundance mRNAs in dendrites in vivo (16, 17), probably due to lack of sensitivity and/or compromised tissue morphology. Our modified protocol has been used to detect the well-known dendritic mRNA CaMKIIα, but also less abundant dendritic mRNAs, such as PSD-95 and GluR1 (17). Here, we show seizure-induced dendritic localization of Arc/Arg3.1 mRNA and dendritic localization of microtubule associated protein 1b (MAP1B) mRNA in vivo (Fig. 3).

Fig. 3.

(a, b) FISH detection of Arc/Arg3.1 mRNA in dendrites of the mouse dentate gyrus under control conditions (a) and 4 h following kainic acid-induced seizure (b). Note that the FISH method using RNA probes and detection with fluorescein-tyramide amplification is suitable to detect Arc/Arg3.1 mRNA specific signal in distal dendrites of single cells under control conditions ((a), indicated by arrows), as well as in the entire granule cell (gc) and molecular layer (ml) following induction of synaptic activity (b). (c, d) FISH detection of MAP1B mRNA in dendrites of the mouse cortex co-immunostained for the dendritic marker protein MAP2. (c) Overlay of the MAP1B FISH signal (green) with MAP2 protein (red  ) staining suggests that MAP1B mRNA is targeted into dendrites of the cortex (upper panel  ). A MAP1B sense probe does not detect any specific signal (lower panels). (d) Magnification of a single cortical neuron (white box in (c) indicates the area shown in (d)) demonstrates that MAP1B mRNA positive granules can be visualized in distal regions (≥50 μm) of a dendrite.

Several laboratories have recently developed improved FISH protocols that allow for analysis of mRNA localization (18). We find the FISH techniques presented here to be especially powerful to detect and analyze high- and low-abundance mRNA species abundant mRNAs in fine neuronal structures in vitro and in vivo.

2 Materials

If not noted otherwise, all solutions are prepared on the day of the experiment with water (H₂O) taken freshly from a Milli-Q Synthesis purification system (or similar), which produces pyrogen- and nuclease-free H₂O with a total organic carbon content of 2–5 ppb, and pyrogen levels (EU/ml) of <0.001. If these guidelines are followed, DEPC-treatment, autoclaving, or filtration is not necessary (if not otherwise noted). All solutions are prepared and stored in sterile, RNAse-free microcentrifuge and conical tubes, or autoclaved glassware. Unless otherwise stated, all chemicals are purchased from Sigma-Aldrich, St. Louis, MO.

2.1 Materials for Fluorescence In Situ Hybridization on Cultured Neurons

2.1.1 Hippocampal Neuron Culture

The materials needed for the hippocampal neuron dissection and culture including dissection tools, media and culture vessels, have been described previously (19). All experiments herein use hippocampal neurons harvested from E18 rat embryos.

2.1.2 Digoxigenin-Labeled Oligonucleotide Probe Preparation and Labeling

• 1. Oligonucleotide probes synthesized with internal T(C6)-amino 5′ modifications and purified by reverse-phase HPLC (Biosearch Technologies, Novato, CA).

• 2. Digoxigenin-3-O-methylcarbonyl-ε-aminocaproic acid-N-hydroxysuccinimide ester (Digoxigenin-NHS; Roche Applied Science, Indianapolis, IN).

• 3. Dimethyl formamide (DMF).

• 4. 0.1 M sodium borate buffer at pH 8.8.

• 5. Sephadex G50 gel filtration columns (GE Healthcare Biosciences, Pittsburgh, PA).

• 6. 70% ethanol.

• 7. Sodium acetate solution: 3 M NaC₂H₃O₂ in H₂O, adjust to pH 5.2 using acetic acid.

• 8. DIG Nucleic Acid Detection Kit (Roche Applied Science).

• 9. Zeta-Probe blotting membrane (BioRad, Richmond, CA).

2.1.3 Fluorophore-Labeled Oligonucleotide Probe Preparation and Labeling

• 1. Oligonucleotide probes synthesized with internal T(C6)-amino 5′ modifications and purified by reverse-phase HPLC (Biosearch Technologies).

• 2. Amersham CyDye mono-reactive dye pack (Cy3 or Cy5; GE Healthcare Biosciences) or AlexaFluor 488 Amine Labeling Kit (Invitrogen, Carlsbad, CA).

• 3. 0.1 M sodium carbonate buffer at pH 8.8.

• 4. Materials listed in Subheading 2.1.2, items 5–7 are also needed.

2.1.4 Fluorescence In Situ Hybridization with Digoxigenin-Labeled Oligonucleotide Probes

• 1. Sterile 12-well plates, nuclease-free (BD Biosciences, San Jose, CA).

• 2. 4% paraformaldehyde in 0.1 M phosphate buffer (PB) with 5 mM MgCl₂ at pH 7.4 (see Note 1 for preparation).

• 3. 10× phosphate buffered saline (PBS: 0.01 M KH₂PO₄, 0.1 M Na₂HPO₄, 1.37 M NaCl, and 0.027 M KCl; Roche Applied Science).

• 4. 1× PBS with 5 mM MgCl₂.

• 5. Salmon sperm DNA (10 mg/ml, Invitrogen, store at −20°C).

• 6. E. coli tRNA solution: 10 mg/ml tRNA (Roche Applied Science) in H₂O. 500 μl aliquots are stored at −20°C for up to a year.

• 7. 20× sodium citrate buffer (SSC): 3.0 M NaCl and 0.3 M sodium citrate at pH 7.0 (Roche Applied Science).

• 8. Deionized formamide, store at 4°C.

• 9. Dextran sulfate solution: 50 mg/ml dextran sulfate in H₂O (see Note 2 for preparation), 1 ml aliquots can be stored at −20°C for up to a year.

• 10. Hybridization Buffer (HB): 200 μl dextran sulfate (50 mg/ml), 200 μl bovine serum albumin (BSA; 20 mg/ml; Roche Applied Science; store at −20°C), 100 μl ribonucleoside vanadyl complexes (RVC; 200 mM in H₂O; aliquoted and stored at −20°C), 100 μl 20× SSC buffer, 10 μl 10 mM PB, and 390 μl H₂O. The solution is mixed by vortexing. HB is made fresh immediately before use and kept on ice.

• 11. Tris-buffered saline (TBS): 50 mM Tris–HCl and 150 mM NaCl at pH 7.4.

• 12. TBS with 0.3% Triton (v/v).

• 13. TBS with 0.1% Triton (v/v).

• 14. Tris/Glycine buffer: 200 mM Tris-HCl at pH 7.4 and 0.75% glycine (w/v).

• 15. Blocking buffer: 2% BSA Fraction V (w/v; Roche Applied Science) and 2% FBS (v/v) in TBS with 0.1% Triton.

• 16. Immunofluorescence (IF) buffer: 1% BSA Fraction V (w/v) and 1% FBS (v/v) in TBS with 0.1% Triton.

• 17. Mouse anti-digoxigenin antibody (Jackson Immunoresearch Laboratories, West Grove, PA).

• 18. Donkey anti-mouse Cy3-conjugated antibody (optionally Cy2- or Cy5-conjugated antibodies can be used; Jackson Immunoresearch Laboratories).

• 19. 1× PBS.

• 20. DAPI (4′,6-diamidino-2-phenylindole).

• 21. Mounting media (see Note 3 for preparation).

• 22. Propyl gallate.

• 23. Superfrost glass microscope slides (Fisher Scientific, Fair Lawn, NJ).

2.1.5 Fluorescence In Situ Hybridization with Fluorophore-Labeled Oligonucleotide Probes

• 1. The materials listed in Subheading 2.1.4, items 1–10 and 19–23 are necessary to complete FISH with fluorophore-labeled oligonucleotide probes.

• 2. Optionally, if protein immunocytochemistry is to be conducted in addition to fluorophore-labeled oligonucleotide mRNA detection, then the materials listed in Subheading 2.1.4, items 1–23 are necessary.

2.2 Materials for Fluorescence In Situ Hybridization on Brain Tissue

2.2.1 Preparation of Tissue

• 1. Physiological saline: 0.9% (w/v) NaCl in H₂O, supplemented with ≥1 USP unit/ml heparin sulfate.

• 2. 4% paraformaldehyde: 4% paraformaldehyde, 0.2 M NaH₂PO₄, and 1 mM MgCl₂ at pH 7.4; filtrated through grade 1 cellulose filters (GE Healthcare Biosciences, see Note 4 for preparation).

• 3. 30% (w/v) sucrose: 30 g sucrose (nuclease-free) in 1× PBS (1:10 dilution of 10× PBS (Roche Applied Science).

• 4. Tissue-Tek OCT media (Sakura Finetek, Torrance, CA).

• 5. Superfrost Plus microscope slides (Fisher Scientific).

• 6. Cryostat/Microtome (suitable to cut 10–15 μm thick frozen sections).

2.2.2 Preparation of Riboprobes

• 1. Plasmid containing the cDNA of interest flanked by Sp6, T7 or T3 RNA polymerase promoters and unique restriction sites (e.g., pcDNA3 from Invitrogen).

• 2. Restriction enzymes and buffers from any manufacturer (e.g., Fermentas, Glen Burnie, MD), phenol/chloroform/isoamyl (25:24:1) solution, chloroform, and ethanol.

• 3. DIG RNA labeling Kit (Sp6/T7) (Roche Applied Science).

• 4. tRNA solution (see Subheading 2.1.4, item 6).

• 5. 3 M sodium acetate: 3 M NaC₂H₃O₂ at pH 5.2 (pH adjusted with acetic acid).

• 6. 0.1 M DTT dissolved in H₂O.

• 7. 0.4 M NaHCO₃ dissolved in H₂O.

• 8. 0.6 M Na₂CO₃ dissolved in H₂O.

• 9. Neutralization solution: 3 M NaC₂H₃O₂ at pH 6 in H₂O (pH adjusted with acetic acid).

• 10. Glycogen solution: 20 mg/ml glycogen in H₂O (Roche Applied Science).

2.2.3 Pretreatment of Brain Sections, Hybridization, and Washes

• 1. Glass staining dish with cover (Fisher Scientific).

• 2. SSC solutions: dilutions of 20× SSC stock solution (Roche Applied Science) with H₂O.

• 3. 0.1 M Triethanolamine-HCl: 0.1 M triethanolamine in H₂O at pH 8.0 (pH adjusted with HCl).

• 4. Acetic anhydride (Fisher Scientific).

• 5. Methanol/Acetone solution: 50% (v/v) Methanol and 50% (v/v) Acetone.

• 6. Hybridization buffer: 4× SSC, 50% (v/v) formamide, 1× Denhardts (50× Denhardt’s from Invitrogen, contains 1% (w/v) Ficoll (type 400), 1% (w/v) polyvinylpyrrolidone, and 1% (w/v) bovine serum albumin), 10% (w/v) dextran sulfate (for preparation of dextran sulfate solution, see Subheading 2.1.4, item 9), 0.5 μg/ml herring sperm ssDNA (10 mg/ml solution from Roche Applied Science), and 0.25 μg/ml tRNA from E. coli, slowly mixed on a rotator to avoid air bubbles, stable at −80°C for 2 weeks.

• 7. Immunostain moisture chamber, black (Evergreen Scientific, Los Angeles, CA).

• 8. HybriSlip hybridization covers (Invitrogen).

• 9. RNAse A solution: 10 mg/ml RNAse A (Ribonuclease A from bovine pancreas) in H2O, store in 0.5 ml aliquots at −20°C.

2.2.4 Detection

• 1. 3% H₂O₂ solution: 1:10 dilution of a 30% H₂O₂ solution (Fisher Scientific) in 1× SSC.

• 2. 10× TBS 100: 1.5 M NaCl and 1 M Tris–HCl at pH 7.5; autoclave, stable at room temperature for up to a month.

• 3. TBS 100 buffer (tris buffered saline): 1:10 dilution of 10× TBS.

• 4. TNB buffer: 0.5% (w/v) blocking reagent in 1× TBS 100, prepared from 10× TBS and 10% (w/v) blocking reagent (Roche Applied Sciences, prepared as described in the manufacturer’s instructions).

• 5. TNT buffer: 1× TBS 100 with 0.5% (v/v) Tween 20.

• 6. Tyramide signal amplification (TSA) system, fluorophore-coupled (PerkinElmer, Waltham, MA).

• 7. Mounting media (see Subheading 2.1.4, items 21 and 22).

3 Methods

3.1 Methods for Fluorescence In Situ Hybridization on Cultured Neurons

The FISH methods described herein use digoxigenin-labeled oligonucleotide probes detected by fluorophore-conjugated antibodies (Fig. 1) and fluorophore-labeled oligonucleotide probes (Fig. 2) to localize mRNA molecules within subcellular compartments of cultured neurons.

3.1.1 Hippocampal Neuron Culture

Hippocampal neuron cultures are prepared as described previously (19). Briefly, hippocampal neurons are dissociated and plated at low density on 15 mm poly-l-lysine-coated glass coverslips (No.1, Carolina Biological Supply, Burlington, NC). Neurons are cocultured with glial feeder layers in Neurobasal media supplemented with Glutamax and B27 (Invitrogen) at 37°C with 5% CO₂. For mRNA detection in mature dendrites and spines, neurons are cultured for 14–21 days in vitro (DIV). To detect axonal mRNAs, neurons are cultured for 3–5 DIV.

3.1.2 Digoxigenin-Labeled Oligonucleotide Probe Preparation and Labeling

• 1. Oligonucleotide sequence design: Antisense oligonucleotide probes are designed to be approximately 50 nucleotides in length with 45–55% GC content, complementary to unique, non-overlapping target mRNA sequences, and with as little homology to other sequences as possible. Selected sequences should have minimal secondary structure and favorable hybridization properties assessed by software (OLIGO, Molecular Biology Insights, Inc., Cascade, CO). To allow precise control of hapten incorporation, five thymidine residues approximately ten bases apart are amino-modified. The amino-modification allows the direct chemical conjugation of succinimide ester compounds (e.g., digoxigenin-NHS or fluorophore-NHS) to the oligonucleotide. To increase signal detection, four distinct oligonucleotides are produced for each target mRNA. As a negative control, an oligonucleotide with a scrambled antisense sequence or the sense sequence is prepared for each target mRNA. The probes are purified by reverse-phase HPLC and resuspended in H₂O at 5–10 μg/μl.

• 2. 20 μg of the oligonucleotide(s) are dried in a speed vacuum, and then resuspended in 200 μl 0.1 M sodium borate buffer. If four oligonucleotides are to be used to detect a single mRNA, then 5 μg of each oligonucleotide are used. The digoxigenin-NHS is dissolved in DMF at 1.67 mg/ml, and 200 μl digoxigenin solution is added to 200 μl of the oligonucleotide suspension. The probe solution is rotated overnight at room temperature and protected from light.

• 3. Oligonucleotide purification: A G50 Sephadex column is used for oligonucleotide purification. The labeled oligonucleotides are concentrated using a speed vacuum and resuspended in 0.1 M sodium carbonate buffer. The oligonucleotide suspension volume must be 100 μl or less and at a concentration no greater than 1 μg/μl to be applied to the column. The column preparation, oligonucleotide application, and fraction collection are completed as per the manufacturer’s instructions (GE Healthcare). The oligonucleotides are precipitated from each fraction by adding 100% ethanol (2.5 times the fraction volume) and 3 M sodium acetate (0.1 times the fraction volume). The samples are placed at −20°C for at least 2 h, then centrifuged in a microcentrifuge at 20,000  ×  g for 20 min at 4°C. The DNA pellets are rinsed with 70% ethanol and centrifuged at 20,000  ×  g for 10 min at 4°C, twice. The DNA is resuspended in H₂O at approximately 100 μg/ml, assuming full extraction of DNA from the column (i.e., if 20 μg of DNA was used for labeling, dissolve DNA in a total of 200 μl H₂O).

• 4. The efficiency of digoxigenin labeling is determined by an alkaline phosphatase immunoassay using the DIG Nucleic Acid Detection Kit. 1 μl of each fraction is dotted on a Zeta-Probe blotting membrane in duplicate, allowed to fully dry, and then crosslinked to the membrane using UV light. The blot is then processed for digoxigenin detection as per the manufacturer’s instructions (Roche Applied Science). The reaction is quenched by washing with H₂O and the membrane is allowed to dry at room temperature (see Note 5).

3.1.3 Fluorophore-Labeled Oligonucleotide Probe Preparation and Labeling

• 1. The oligonucleotide probes are designed as described in Subheading 3.1.2, step 1.

• 2. Oligonucleotide preparation and labeling: 20 μg of oligonucleotide(s) are dried using a speed vacuum and resuspended in 70 μl 0.1 M sodium carbonate buffer. Dissolve 1 vial (GE Healthcare Cy3 or Cy5) or 1 mg (Alexa 488) of dye in 30 μl of DMSO. The dye and oligonucleotide solutions are combined, left at room temperature overnight with occasional vortexing, and protected from light.

• 3. The fluorophore-labeled oligonucleotides are purified and precipitated as described in Subheading 3.1.2, step 3.

• 4. The specific activity of fluorophore-labeled oligonucleotides is determined using a spectrophotometer. The resuspended oligonucleotide is diluted 1:100 in H₂O. The DNA concentration is measured by absorbance at 260 nm, and the fluorescence is determined by measuring the absorbance at the maximum absorption wavelength for the fluorophore used (e.g., 488 nm for AlexaFluor 488, 514 nm for Cy3, and 643 nm for Cy5). The oligonucleotide probe is resuspended at a final concentration of 25 ng/μl.

3.1.4 Fluorescence In Situ Hybridization with Digoxigenin-Labeled Oligonucleotide Probes

• 1. Hippocampal neurons cultured on glass coverslips are placed into 12-well plates containing 4% paraformaldehyde, and the neurons are fixed for 20 min at room temperature. The plate should not be moved during fixation as this can compromise the fixation process.

• 2. The neurons are washed three times in 1× PBS/MgCl₂ for 5 min. Unless otherwise noted, all washes are completed at room temperature on an orbital shaker.

• 3. To equilibrate the samples for hybridization, the neurons are washed in 1× SSC buffer for 10 min, and then in 1× SSC with 40% formamide for 5 min.

• 4. The HB is prepared while the neurons are being fixed, washed, and equilibrated. Approximately 30 μl of HB is needed for each coverslip (see Note 6).

• 5. Prehybridization: To minimize nonspecific oligonucleotide hybridization, the neurons are incubated with prehybridization solution for 1.5 h at 37°C. For each coverslip, 20 μg of salmon sperm DNA and 20 μg of tRNA are dried using a speed vacuum, then resuspended in 15 μl 2× SSC with 80% formamide, heated for 5 min at 95°C, and briefly cooled on ice. 15 μl of HB is added to the formamide mixture, per coverslip, and mixed well. A moisture chamber is assembled with a piece of parafilm laid flat where the coverslips will be placed for incubation. 28 μl prehybridization solution is dotted on the parafilm for each coverslip and the coverslips are placed on the solution neuron-side down using fine forceps. The chamber is covered and placed at 37°C for 1.5 h.

• 6. Hybridization: The probe solution for one coverslip contains 25 ng of labeled oligonucleotide probes, 20 μg of salmon sperm DNA, and 20 μg of tRNA (final probe concentration 0.5–1.0 ng/μl). The oligonucleotide and carrier molecules are dried in speed vacuum, resuspended in 15 μl of 2× SSC with 80% formamide, heated at 95°C for 5 min, and briefly cooled on ice. 15 μl of HB is added to the formamide mixture, per coverslip, and mixed well. On a new piece of parafilm, 28 μl of probe solution is dotted for each coverslip. The coverslips are carefully lifted off of the parafilm after prehybridization using fine forceps and blotted with a laboratory tissue to remove excess prehybridization solution (see Note 7). The coverslips are immediately placed onto the hybridization solution and put at 37°C for 5 h.

• 7. The coverslips are carefully removed from the parafilm and the neurons are washed two times with prewarmed 1× SSC with 40% formamide for 20 min at 37°C. Then, the neurons are washed briefly three times with 1× SSC, followed by two 5 min washes in 1× SSC.

• 8. The neurons are equilibrated in 1× PBS/MgCl₂ for 5 min, and then, post-fixed in 4% paraformaldehyde for 5 min at room temperature. The neurons are then washed three times in 1× PBS/MgCl₂ for 5 min.

• 9. The neurons are equilibrated in 1× TBS for 10 min at room temperature.

• 10. Then, the neurons are permeabilized with 0.3% Triton-TBS for 10 min, washed with Tris-Glycine buffer for 5 min, and washed with 0.1% Triton-TBS for 5 min.

• 11. The neurons are incubated in Blocking buffer for 1 h at room temperature.

• 12. The neurons are washed in immunofluorescence (IF) buffer for 5 min, and then incubated with mouse anti-digoxigenin antibody diluted 1:1,500 in IF buffer for 1 h at room temperature (if double labeling is desired, then additional primary antibodies can be added at this step). The antibody incubations are completed in a moisture chamber. 30 μl of antibody solution is dropped onto a flat piece of parafilm and the coverslips are carefully inverted onto the antibody solution with fine forceps.

• 13. The neurons are washed three times for 10 min in 2 ml of IF buffer.

• 14. The neurons are incubated with donkey anti-mouse fluorophore-conjugated antibody diluted 1:1,000 in IF buffer for 30 min at room temperature.

• 15. The neurons are washed three times for 10 min in 2 ml of IF buffer.

• 16. If DAPI stain is desired, then the neurons are washed in 1× PBS for 5 min, treated with DAPI (1:1,000 in 1× PBS) for 5 min, and washed for 5 min with 1× PBS.

• 17. Mounting media aliquots are thawed 2–3 h before the coverslips will be ready for mounting. Propyl gallate (0.6 mg/ml) is added to the mounting media, and the mounting media is rotated at room temperature for at least 1 h in the dark. The mounting media is centrifuged at 20,000  ×  g for 5 min to pellet non-dissolved propyl gallate.

• 18. The coverslips are rinsed briefly with H₂O to remove salts and detergents, air-dried, and then mounted on glass slides with 15 μl of mounting media.

• 19. The slides are dried at room temperature overnight in the dark, and then stored at −20°C.

• 20. For high-resolution images, neurons are visualized on a widefield fluorescence microscope (Nikon Eclipse TE300 inverted microscope or similar). Images are captured with a cooled CCD camera (Quantix; Photometrics, Tuscon, AZ, or similar), and then deconvolved using a 3D blind algorithm (AutoQuant X; Cybermetrics, Phoenix, AZ, or similar).

3.1.5 Fluorescence In Situ Hybridization with Fluorophore-Labeled Oligonucleotide Probes

• 1. In vitro FISH with fluorophore-labeled oligonucleotides is conducted exactly as described in Subheading 3.1.4, steps 1–8.

• 2. Optionally, if protein immunocytochemistry and fluorophore-labeled oligonucleotide FISH are to be conducted, then the protocol described in Subheading 3.1.4, steps 1–20 are completed.

3.2 Methods for Fluorescence In Situ Hybridization in Brain Tissue

The here presented FISH method for brain tissue sections using digoxigenin-labeled RNA probes (riboprobes) and fluorophore-coupled tyramide signal amplification provides preservation and accessibility of the tissue, and allows for detection of high- and low-abundance mRNAs in dendrites in vivo ((17), also see Fig. 3).

3.2.1 Preparation of Tissue Sections

• 1. To protect tissue morphology and mRNA structure, mice at postnatal day 21 are deeply anesthetized with an inhalative anesthetic (e.g., isoflurane) and transcardially perfused (ca. 5 ml/min) with 80 ml of prewarmed (37°C) physiological saline supplemented with 1 unit/ml of heparin sulfate to reduce blood clotting. Then, mice are perfused with 120 ml of prewarmed (37°C) 4% paraformaldehyde (see Note 8).

• 2. The brain is removed from the skull and stored in 5 ml of 4% paraformaldehyde at 4°C overnight.

• 3. After 16–18 h, the tissue is placed in 10 ml of 30% (w/v) sucrose in 1× PBS and stored at 4°C for 24 h.

• 4. Brains are placed in Tissue-Tek and rapidly frozen using liquid nitrogen. Frozen brains are wrapped in aluminum foil and stored at −80°C.

• 5. At the day of experiment, desired brain regions are cut in 10–15 μm thick sections using a cryostat/microtome and mounted on Superfrost Plus microscope slides (see Note 9). Mounted sections can be stored at −80°C for several weeks to months, but best results are usually obtained if freshly cut sections are immediately processed for FISH.

3.2.2 Preparation of Riboprobes

• 1. For the design of efficient riboprobes, the following guidelines should be followed: The cDNA used for in vitro transcription of the riboprobe contains 1.2–1.8 kb of the target sequence with as little homology to other sequences as possible to provide sufficient sequence specificity (see Note 10). The cDNA is subcloned into a plasmid containing promoters for two different RNA polymerases (Sp6, T7 or T3). Plasmids require two unique restriction sites, one at the 5′-, and one at the 3′-end of the sequence to allow for linearization prior to in vitro transcription. Promoters for two different polymerases, as well as single restriction sites on the 5′- and 3′-end of the cDNA allows for transcription of an antisense and a sense riboprobe from the same cDNA construct. Sense riboprobes are used as control for nonspecific signal and background labeling.

• 2. To prepare cDNA for in vitro transcription, 20 μg cDNA-containing plasmid is linearized with the appropriate restriction enzyme according to the manufacturer’s protocol (see Note 11). 2% of the reaction volume should be checked for complete linearization by DNA gel electrophoresis. If linearization is complete, the plasmids are purified by phenol/chloroform extraction and precipitated with ethanol following standard protocols. DNA precipitates are dissolved in 15 μl of H₂O and stored at −20°C.

• 3. Digoxigenin-labeled transcripts are synthesized using the DIG RNA labeling Kit (Sp6/T7) from Roche Applied Science according to the manufacturer’s protocol with 2 μl (1–2 μg) of the DNA obtained in Subheading 3.2.2, step 2. If necessary, Sp6 or T7 polymerase can be substituted with T3 polymerase (not contained in the kit, but available from Roche Applied Science). DNase treatment (optional in the original protocol) should be performed. The in vitro transcribed RNA probes contain ca. 4–5% UTPs labeled with digoxigenin (see Note 12). Riboprobes are precipitated by adding 80 μl H₂O, 10 μl tRNA solution, 20 μl 3 M sodium acetate (pH 5.2), and 300 μl ethanol. Samples are briefly vortexed and incubated at −20°C for at least 1 h. Riboprobes are centrifuged at 20,000  ×  g for 20 min at 4°C, the supernatant is removed, and 1 ml 75% Ethanol is added. Samples are mixed thoroughly, and centrifuged at 20,000  ×  g for 5 min at 4°C. After removal of the supernatant, the precipitated RNA is air-dried for 20 min at 37°C (see Note 13).

• 4. To provide sufficient tissue penetration, size reduction by alkaline hydrolysis is performed (20). Riboprobes are resuspended in 160 μl of 0.1 M DTT. 20 μl each of 0.4 M NaHCO₃ and 0.6 M Na₂CO₃ are added, the samples are mixed and incubated at 60°C. The incubation time depends on the size of the original transcript (L ₀) and the desired final length (L _F, 0.1 kb is a good starting length), and can be ­calculated with this formula: t[min]  =  (L ₀[kb]  −  L _F[kb])/L ₀[kb]  ×  L _F[kb] (see Note 14). Following size reduction, samples are immediately put on ice, neutralized with 7 μl ice-cold neutralization solution and precipitated by adding 2 μl glycogen solution and 500 μl ice-cold ethanol. Samples are mixed thoroughly and incubated on ice for at least 30 min, followed by centrifugation at 20,000  ×  g for 20 min at 4°C. Precipitates are washed with 75% ethanol as described in Subheading 3.2.2, step 3. Air-dried RNA is resuspended in 20 μl H₂O, diluted with 50 μl hybridization buffer and mixed thoroughly. Riboprobes are stored at −80°C. Riboprobes in hybridization buffer are stable at −80°C for about 2 months and do not lose activity following repeated freeze and thaw cycles (up to five times).

3.2.3 Pretreatment of Brain Sections

• 1. Optional: Slides with mounted brain sections are removed from the −80°C storage and brought to room temperature. Fresh and frozen slides are air-dried at room temperature.

• 2. Dried slides are placed upright into a slotted glass jar. Unless otherwise noted, incubation and washing steps are conducted at room temperature on an orbital shaker. All solutions and buffers can be poured slowly in or out of the glass jar.

• 3. Sections are postfixed for 5 min in ice-cold 4% paraformaldehyde.

• 4. Sections are washed twice for 10 min with ice-cold 2× SSC.

• 5. To reduce nonspecific hybridization, free amino-groups are acetylated with acetic anhydride. Sections are equilibrated for 5 min in ice-cold 0.1 M triethanolamine-HCl, pH 8.0. Immediately before acetylation, 50 ml ice-cold 0.1 M triethanolamine-HCl at pH 8.0 is placed in a sealable tube and the pre-equilibration solution is poured off the sections. 750 μl of acetic anhydride is added to the tube containing 0.1 M triethanolamine-HCl at pH 8.0, mixed thoroughly for 2–3 s, immediately poured on the slides and incubated for 10 min.

• 6. Sections are washed briefly three times with ice-cold H₂O and incubated for 5 min in ice-cold Methanol/Acetone solution.

• 7. Sections are washed twice for 10 min in ice-cold 2× SSC.

• 8. For prehybridization, slides are placed upright outside the jar for 2–3 min to remove excess liquid, and then placed horizontally in a moisture chamber, humidified with paper tissues soaked with 50% formamide in 2× SSC. Sections are covered with 400 μl hybridization buffer and incubated in the sealed moisture chamber for at least 2 h at 55°C. Prehybridization can be extended to up to 24 h.

3.2.4 Hybridization

• 1. Antisense and sense riboprobes are diluted 1:5, 1:10, and 1:20 in hybridization buffer and mixed thoroughly (see Note 15). 50 μl of hybridization buffer per brain section is needed. To reduce secondary structures of riboprobes, the hybridization solutions are incubated at 95–99°C for 5 min and immediately placed on ice for 5 min.

• 2. Prehybridized sections are removed from the moisture chamber and placed upright on a paper tissue to allow prehybridization solution to drain off. Excess solution can be removed carefully using a tissue without touching the sections.

• 3. Slides are placed back in the moisture chamber; 50 μl of hybridization solution is applied carefully directly onto the sections and covered with HybriSlip hybridization covers.

• 4. Sections are incubated for 14–18 h at 55°C (see Note 16). The probe is hybridized in a solution containing final concentrations of 50% formamide and 4× SSC.

3.2.5 Post-hybridization Washes

• 1. Following hybridization, slides are placed in 2× SSC for 10 min to remove HybriSlip hybridization covers (see Note 17).

• 2. After removal of the hybridization covers, sections are washed again for 10 min in 2× SSC.

• 3. To remove non-hybridized RNA, sections are incubated for 15 min at 37°C with 10 μg/ml RNAse A in 2× SSC ­(pre-warmed to 37°C).

• 4. Sections are washed twice for 10 min at room temperature in 2× SSC.

• 5. Sections are equilibrated for 5 min in 0.5× SSC at room temperature, followed by 30 min incubation in 0.5× SSC at 50°C. The 0.5× SSC solution should be prewarmed to 50°C prior usage.

• 6. Slides are washed twice for 10 min in 2× SSC at room temperature.

3.2.6 Detection

• 1. To inactivate endogenous peroxidases, sections are incubated for 15 min in 3% (v/v) H₂O₂ in 1× SSC, followed by three times 5 min washes in 1× SSC.

• 2. Sections are then incubated for 5 min in TBS 100 buffer.

• 3. Nonspecific binding sites are blocked by 30 min incubation at room temperature in TNB buffer.

• 4. After blocking, slides are placed upright on a paper tissue for 1–2 min to allow excess TNB buffer to drain off. Antibody solution is prepared by diluting sheep anti-digoxigenin-POD, Fab fragments (Roche Applied Science) in TNB buffer (starting dilution 1:1,000, needs to be optimized for each lot of antibody individually, see Note 18). Slides are placed in a sealable moisture chamber humidified with wet paper tissues, covered with approximately 100–200 μl of antibody solution each and incubated for 2 h at room temperature (see Note 19). Optional: If codetection for a specific protein is desired, the first antibody specific to the respective protein can be included here.

• 5. Sections are washed five times 10 min in TNT buffer. [Optional: If simultaneous immunohistochemistry is performed, a 1 h incubation in secondary antibody (1:200 in TNB buffer, reactive to the species the first antibody was generated in and coupled to a fluorophore other than the one used in the following tyramide signal amplification) can be included, followed by five times 10 min washes in TNT buffer. Incubation in secondary antibody, as well as washes and all following steps should be performed in the dark.]

• 6. The tyramide signal amplification (21) is performed in a horizontal moisture chamber, protected from light. Slides are placed upright on a paper tissue for 1–2 min to allow TNT buffer to drain off. Slides are then placed in the horizontal chamber and sections are covered with 50 μl fluorophore-coupled tyramide working solution (PerkinElmer, tyramide reagent diluted 1:50 in amplification buffer, according to the manufacturer’s instructions). Sections are incubated in TSA for exactly 8 min and slides are placed immediately in TNT buffer (see Note 20).

• 7. Sections are washed five times 10 min in TNT buffer.

• 8. Slides are briefly washed in H₂O to remove salts and detergents, air-dried, mounted with glass coverslips using mounting media as described in Subheading 3.1.4, step 17. After drying overnight at room temperature in the dark, microscope slides can be stored at −20°C.

• 9. For high-resolution visualization, tissue sections are imaged as z-stacks using a confocal laser scanning microscope.

4 Notes

• 1. To prepare 4% paraformaldehyde in 0.1 M PB with 5 mM MgCl₂, paraformaldehyde powder is added to 0.1 M PB, then heated and stirred. Sodium hydroxide is added to aid dissolution of paraformaldehyde. The pH is brought to 7.4, and then 1 M MgCl₂ is added to a final concentration of 5 mM MgCl₂. The solution is brought to the final volume with 0.1 M PB, and filtered.

• 2. Due to poor solubility of dextran sulfate in H₂O, the 50 mg/ml solution is heated to 80°C and inverted frequently (ca. 1 h). Insoluble particles are pelleted by centrifugation at 1,000  ×  g for 10 min at 4°C, then the solution is brought to the final volume with H₂O.

• 3. To prepare mounting medium, 25 g polyvinyl alcohol is added to 100 ml 1× PBS slowly while stirring and protected from light. The stirring is continued overnight. After the polyvinyl alcohol is dissolved, the pH is adjusted to 7.2 using pH indicator strips as the solution is very viscous. 50 ml glycerol is added and the solution is stirred overnight, again. Once the solution is homogenous, it is aliquoted and stored at −20°C.

• 4. Due to poor solubility of PFA in H₂O, 4% PFA solution should be prepared using the following protocol. To prepare 250 ml, 1.1 g NaOH and 10 g paraformaldehyde are added to 200 ml H₂O and stirred until dissolved (ca. 10 min, residual undissolved solids may still be visible). 6 g anhydrous NaH₂PO₄ is added (final concentration 0.2 M) and the solution is stirred until dissolved, followed by 250 μl of 1 M MgCl₂ (final concentration 1 mM). The pH is adjusted to 7.4 using 10 N NaOH, and the volume is raised to 250 ml with H₂O, followed by filtration through grade 1 cellulose filters.

• 5. Dots appear quickly and will saturate. So, samples are checked every 30 s after addition of detection buffer and the reaction is quenched with water prior to saturation. Make your judgment or image the blot prior to drying in order to assess which fractions contain efficiently labeled probes.

• 6. Approximately 10–15% extra HB is prepared for each experiment because the HB solution is viscous due to the high concentration of dextran sulfate, and an appreciable amount will be lost during pipetting. Oligonucleotide probes designed to specific mRNA sequences have approximately 50% GC content, whereas an oligonucleotide probe to detect poly(A) mRNA is composed of only thymidine bases. Given that probes rich in G-C base pairs have higher melting temperatures than those rich in A-T pairs, the hybridization buffer stringency must be reduced. In all washes and hybridization buffers, 15% formamide is used instead of 40% and the length of hybridization can be shortened to 3 h at 37°C.

• 7. High surface tension holds the coverslips to the parafilm. It may be helpful to put a drop of 1× SSC buffer on the ­coverslip to aid in the gentle removal using the fine forceps. The careful removal of coverslips from parafilm is crucial; if you drag the coverslips the neuronal morphology will be comprised.

• 8. To ensure good tissue preservation, efficient perfusion of the mice with physiological saline should be monitored, e.g., by change of the color of the liver from dark red to pale. Fixation can be monitored by checking the stiffness of the mouse body (e.g., the neck or tail).

• 9. Thinner sections (10 μm) are beneficial for detection of dendritic mRNAs, especially those with low abundance. However, if the morphology of the tissue suffers, up to 15 μm thick sections can be used.

• 10. Riboprobes spanning large parts of the desired sequence are preferred, as they ensure sequence specificity and circumvent detection difficulties in vivo, where parts of the sequence might be masked with e.g., associated proteins. However, if homologies to other sequences prevent the use of larger cDNA fragments, or if probes specific to certain parts of the mRNA are required (e.g., 5′- or 3′-UTR or the open reading frame), the size of the cDNA to make the riboprobes might be reduced to 400–500 nt.

• 11. For linearization of plasmids, enzymes without endogenous star activity are preferred. Digestion of plasmids should then be performed with an excess of restriction enzyme for ca. 3 h. If linearization is incomplete, more enzyme (usually 0.5–1 μl) can be added and the incubation time should be increased to 4–5 h.

• 12. Riboprobes can be checked for digoxigenin incorporation with a dot blot as described in the manufacturer’s protocol. To assure the right size of the transcript, 2% of the sample can be analyzed on an agarose formaldehyde RNA gel (22). This is especially important if riboprobes are being synthesized for the first time.

• 13. In our experience, additional purification of the in vitro transcribed riboprobe by gel filtration using Sephadex G50 columns to remove unincorporated nucleotides is not necessary, and should be avoided as any additional handling of the probe might compromise its quality.

• 14. Although size reduction of the riboprobe increases tissue penetration, for some probes it might also increase background staining. In this case, final fragment size could be up to 0.2 kb, or probes can be used without size reduction.

• 15. The optimal dilution of the riboprobe can vary depending on the probe sequence, and needs to be adjusted for every newly generated riboprobe.

• 16. Most riboprobes work at a hybridization temperature of 55°C. If the GC content of a specific probe is very high (>60%) or very low (<40%), hybridization temperature should be increased (for high GC content) or decreased (for low GC content) in 3°C increments.

• 17. Usually, HybriSlip hybridization covers come off very easily during the first wash in 2× SSC. If not, it is generally better to remove hybridization covers by lifting them up using forceps as opposed to “sliding” them off, which might compromise tissue morphology.

• 18. To provide sufficient tissue penetration of the antibody, it is necessary that anti-digoxigenin Fab fragments are used.

• 19. The same moisture chamber can be used for hybridization and antibody incubation. After hybridization, however, the moisture chamber has to be cleaned carefully to remove any residual formamide. For antibody incubation and the tyramide reaction, tissues soaked in H₂O should be used to humidify the chamber.

• 20. Incubation time with the tyramide amplification solution is crucial, all sections should be incubated in the tyramide solution for the exact same time, to allow for comparison between different dilutions of the riboprobe, as well as assess background staining of the sense probe. To reduce background staining or increase signal intensity, incubation time can be adjusted.